Improving image luminosity and contrast variation using hybrid statistical strategy / \c Wan Azani bin Wan Mustafa.
According to the literature review, it can be concluded that the objective of image enhancement is to improve the image quality for analyses, especially for segmentation stage. This study is dedicated to develop an innovative image enhancement technique, especially for contrast and luminosity normal...
Saved in:
| Main Author: | |
|---|---|
| Corporate Author: | |
| Format: | Thesis Book |
| Language: | English |
| Published: |
Perlis, Malaysia
School of Mechatronic Engineering, Universiti Malaysia Perlis
2017
|
| Subjects: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | According to the literature review, it can be concluded that the objective of image enhancement is to improve the image quality for analyses, especially for segmentation stage. This study is dedicated to develop an innovative image enhancement technique, especially for contrast and luminosity normalization to improve the visibility of low-quality digital images caused by various reasons such as high dynamic range scene irradiance, poor contrast due to the very narrow dynamic range of the captured scene, very low illumination (low light conditions), non-uniform illumination or the spectral characteristics of the illumination, and the limited resolution of the imaging devices. In summary, the objectives are ; to develop a direct enhancement method using the statistical information this method based on the statistical information such as mean, variance and standard deviation. The target is to solve the luminosity and contrast problem on the non-uniform image ; to evaluate the efficiency of the proposed method in non-uniform dataset images in order to prove the effectiveness of proposed method, two datasets were applied which are document images and weld defect images. A few image quality assessments such as Signal Noise Ratio (SNR), Peak Signal Noise Ratio (PSNR) and Global Contrast Factor (GCF) were utilized ; to demonstrate the performance of the enhanced image on the segmentation stage. Based on the literature, the enhancement process will influence the result of the segmentation. In this study, a few suitable segmentation such as Otsu method, Niblack 8 method and Bernsen method were selected in order to apply on the proposed image. The increment performance between the original image and proposed image was obtained. |
|---|---|
| Physical Description: | xii, 141 pages colour illustration 30 cm |
| Bibliography: | Includes bibliographical references. |
| ISBN: | X210000000870 |